We revisit the quantum reverse Shannon theorem, a central result in quantum information theory that characterizes the resources needed to simulate quantum channels when entanglement is freely available. We derive a universal additive upper bound on the smoothed max-information in terms of the sandwiched Rényi mutual information. This bound yields tighter single-shot results, eliminates the need for the post-selection technique, and leads to a conceptually simpler proof of the quantum reverse Shannon theorem. By consolidating and streamlining earlier approaches, our result provides a clearer and more direct understanding of the resource costs of simulating quantum channels.